Resolve concrete AI models from .planning/config.json for bGSD agents.
npx skills add https://github.com/gonz0w/bgsd-oc --skill model-profiles
Please help me install this Agent Skill. Skill: model-profiles Source: https://github.com/gonz0w/bgsd-oc/tree/main/skills/model-profiles Command: npx skills add https://github.com/gonz0w/bgsd-oc --skill model-profiles
Aligns AI capabilities with project needs by centralizing model selection into a reusable contract, reducing drift and manual configuration across agents.
Define the model_settings with a default_profile and profile mappings in .planning/config.json, and let the orchestrator resolve concrete models automatically at startup.
High-intent search queries and answers about installing and using this skill.
Centrally manage AI model selection by defining model_settings with a default_profile and profile mappings in .planning/config.json. The orchestrator reads this file and deterministically resolves concrete models at startup, exposing a final contract to all agents.
A centralized model contract maps default profiles like quality, balanced, and budget to specific AI models. It drives consistent agent behavior across orchestration phases without requiring manual model assignments for each individual agent.
Yes, you can apply per-agent overrides to tailor providers without changing the global contract. This enables local experimentation and budget management while maintaining centralized defaults for the rest of your orchestration workflow.
Handle provider changes and budget constraints by mapping different profiles such as quality, balanced, and budget in your config.json. The resolution logic honors both default profiles and per-agent overrides to accommodate your specific financial or provider limitations.
Yes, this configuration-driven resolution operates without external dependencies. It reads your local .planning/config.json file to map profiles to models and exposes the final contract directly to your agents at orchestration startup.
blueif16Create slash command files and initialize plugin structure in Claude plugins.
gustavofsantosAnalyze plan.md to detect project state and route work to the matching Agent Skill.
mapachekurtMonitor Claude conversation token usage and issue warnings at critical thresholds.
bennybennisonRoute model selection to balance quality and cost across task tiers.
